Deep Chest: an artificial intelligence model for multi-disease diagnosis by chest x-rays

Abstract

Background: Artificial intelligence is increasingly being used for analyzing image data in medicine. Objectives: We aimed to develop a computer vision artificial intelligence (AI) application using limited training material to aid in the multi-label, multi-disease diagnosis of chest X-rays. Methods: We trained an EfficientNetB0 pre-trained model, leveraging transfer learning and deep learning techniques. Six thoracic disease categories were defined, and the model was initially trained on images sourced online and chest X-rays from a hospital database for training and internal validation. Subsequently, the model underwent external validation. Results: In constructing and validating Deep Chest, we utilized 453 images, achieving an area under curve (AUC) of 0.98, sensitivity of 0.98, specificity of 0.80, and accuracy of 0.83. Notably, for diagnosing masses or nodules, the sensitivity, specificity, and accuracy were 0.97, 0.81, and 0.83, respectively. We deployed Deep Chest as a free experimental web application. Conclusions: This tool demonstrated high accuracy in diagnosing both single and coexisting pulmonary pathologies, including pulmonary masses or nodules. Deep Chest thus represents a promising AI-based solution for enhancing diagnostic capabilities in thoracic radiology, with the potential to be utilized across various medical disciplines, especially in scenarios where expert support is limited.

Competing Interest Statement

The authors have declared no competing interest.

Funding Statement

This study did not receive any funding

Author Declarations

I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.

Yes

The details of the IRB/oversight body that provided approval or exemption for the research described are given below:

Vast majority of our dataset includes open source chest x-rays, however, additional limited number of patient chest x-rays were used to enrich the dataset. Again, limited retrospective external database of chest x-rays were used for validation purposes. For the limited number of real patients whose chest x-rays were used, due to the retrospective nature, it was not possible to obtain consent from these cases as some of them had unfortunately died at the time of data collection. All chest X-rays used for this study had been de-identified. Thus, no ethical committee approval or consent was obtained.

I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.

Yes

I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).

Yes

I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable.

Yes

Data Availability

All data produced in the present study are available upon reasonable request to the authors

留言 (0)

沒有登入
gif